A&. Radiar. Isor. Vol. 42, No. 11, pp. 1049-1053, Inl. J. Radial. Appl. Instrum. Part A Printed in Great Britain. All rights reserved
1991 Copyright
0883-2889/91 $3.00 + 0.00 0 1991 Pergamon Press plc
An Improved Synthesis of the Inert, Diffusible Blood-flow Tracer, [ ‘sF]Fluoromethane S. JOHN
GATLEY*, RODOLFO FRANCESCHINIt, DAVID J. SCHYLER and ALFRED
Department
of Chemistry,
Brookhaven
National
Laboratory,
RICHARD P. WOLF Upton,
FERRIERI,
NY 11973, U.S.A.
(Received 19 December 1990; in revised form 20 February 1991) An improved procedure for producing [ ‘*F]fluoromethane ([ ‘8F]FM) in batches of several hundred mCi is reported. ‘aF prepared by the InO (p, n) ‘sF reaction in a Hr’*O target is trapped on a small column of Dowex 1 ( x 10) resin to allow recovery of H, ‘*O. One new feature is elution of “F- from the column with 3 mL of CH,CN containing 67 nL_ 1.5 M aqueous (Bu),N+ OH-, after residual H,180 has been removed with dry CH,CN. This “F- solution reacts with CH,I in the nresence of Ae,O directlv to eive [ ‘*F]FM, which is swept out of the reaction vessel with a stream of air, from which CH;i and other vaiors are removed with a C,, SEP-PAK at room temperature. Another new technique is trapping of [ ‘*F]FM on a second SEP-PAK cooled in ethanol/dry ice. After warming the SEP-PAK to room temperature, the trapped [ 18F]FM can be recovered with either H,O or air. The improvements speed the preparation and minimize hands-on operations. The product has no detectable radiochemical impurities, and a specific activity of > 1 Ci/pmol. Non-radioactive CH,CN, CH,I and CH,OH are present at ~0.2 pmol per batch.
Introduction [ “FlFluoromethane
([ 18F]FM) has several advantages for positron emission tomographic (PET) measurement of local cerebral blood flow (LCBF) (Holden et al., 1981, 1983). These include the physical properties of ‘*F. The half-life of 110 min is long enough to permit distribution to PET centers from a regional radiochemical facility, and availability for several hours from a single batch synthesis. The relatively low /I,,, of the “F positron leads to lower tissue radiation doses and possibly better spatial resolution in future tomographs than for other /3+ emitters. In addition, the solubilities of FM in water and in fat (Gatley et al., 1981; Rosenthal and Nickles, 1985), which govern losses in expired air and incorporation into adipose tissue, are such that the agent has a short biological half-life, allowing repeat scanning (30-60min) in a single session. It is also easy to prepare, metabolically inert (Gatley et al., 1981) and can provide values of LCBF without the need for arterial blood sampling (Koeppe et al., 1985b). Fluoromethane may alternatively be radiolabeled with “C (Stone-Elander et al., 1988), and interest has *Author for correspondence. tPresent address: Instituto di Fisiologia Universita
Clinica de1 CNR, di Pisa, Via Savi 8, 56100 Pisa, Italy.
also been expressed (Mulholland et al., 1987) in FM labeled with “F, which has a 65 s half-life. The preparation of [ 18F]FM usually proceeds from [ ‘*F]fluoride which can be prepared by a variety of methods, but most commonly at present by the ‘*O(p, n)‘*F reaction with a water target (Nickles et al., 1986). [‘*F]FM has also been made using cyclotron targetry based on the “Ne(d, ~r)i*F reaction, via the reactive intermediates [ 18F]F2 (Wagner, 1984), N-[ ‘*F]fluoropyridone (Oberdorfer et al., 1988) and [‘*F]diethylaminosulfur trifluoride (Straatman and Welch, 1977). Fluoromethane labeled with either “C or “F has been used in basic medical research (Celesia et al., 1982; Sollevi et al., 1987) clinical applications (Celesia et al., 1984; Lagreze et al., 1987; Lagreze et al., 1988; Levine et al., 1988a,b,c, 1989; Rao et al., 1984) and in PET methodological studies (Heiss et al., 1990; Herholz and Patlak, 1987; Herholz et nl., 1989; Koeppe et al., 1985a,b; Sollevi et al., 1987). The FM method depends on tracer kinetic modeling, on a pixel-by-pixel basis, of the time-course of “F in the brain determined by a temporal sequence of scans (Holden et al., 1981). Because FM is cleared rapidly by exhalation, the biological half-life is quite short; only 2-3% of radioactivity remains in the body after 20 min. Fluoromethane can be given to most subjects by inhalation from an air-bag, followed by a brief
1049
1050
S.
JOHN
GATLEY
(30-45 s) breath hold. Alternatively it can be introduced into a spirometer and administered under a re-breathing protocol. FM can also be easily dissolved in saline and given by i.v. injection. In this case more than half of the administered dose will be expired in the first minute. Previously (Gatley et al., 1981), [18F]FM was obtained by drawing off the gas space through the septum of a sealed reaction vessel into one or more 50 mL syringes. Besides giving a rather large hand dose, this typically resulted in 2&50mL volumes in large syringes which were difficult to shield for transport. There were also concerns about chemical impurities in these preparations. The aims of the present study were to design a system for safely producing high levels (several hundred mCi) of [ ‘*F]FM for human use (either by inhalation or i.v. injection), and to characterize non-radioactive impurities in doses of [ ‘*F]FM.
Experimental Materials Vacutainer”s were obtained from Becton-Dickinson, Rutherford, NJ. Dowex 1 x 10 (200-400 mesh, Cl- form) was obtained from Becker (Phillipsburg, NJ) and converted to the OH -form as described previously (Schlyer et al., 1990). C,,SEP-PAK cartridges were purchased from Waters (Milford, MA). Preparation
of’ F-18 FM
For analysis of impurity vapors product, FM was harvested from the the reaction vessel in a 50mL syringe previously (Gatley et al., 1981), and 1 mL water. The new preparation of [“F]FM is 10 mL Vacutainer initially containing
in crude lBF head-space of as described washed with as follows. A 50mg Ag,O
et al.
and held in an oil bath at 100°C was used as a reaction vessel. The rest of the apparatus (Fig. 1), in addition to a small glass column of Dowex 1, included a 7 mL Vacutainer acting as a cold trap and a SEP-PAK to remove residual vapors. A second SEP-PAK, previously activated with 5 mL ethanol, was immersed in an ethanol/dry ice bath in order to trap [18F]FM. The components were connected with polypropylene tubing and Luer fittings, and 20 ga needles. In some initial experiments the SEP-PAKs were replaced by a 7 mL Vacutainer in which the gas stream was passed through 1 mL of ethanol at various temperatures. With valve A open, irradiated H,lsO containing 18F was pumped from the cyclotron target into the head space above the Dowex column. The Hz ‘“0 was recovered for re-use (syringe 3). Valve A was closed and valve D switched to allow 2 mL of dry CH,CN from syringe I to rinse the column (collected in syringe 4). Valves B and D were than switched to allow introduction of 3 mL of CH,CN containing 67pL of 1.5 M aqueous (Bu),N+ OHinto the reaction vessel. A 1 mL sample of CH,CN containing 100 p L of CH, I were then similarly added (syringe 5), after switching valves B and E appropriately. The reaction mixture was heated for lOmin, and then 50mL air was drawn through the system (syringe 7) at a rate which caused the solvents to boil gently. Trapped [ ‘*F]FM was recovered, after allowing SEP-PAK 2 to warm to room temperature, by switching valve C to allow 5 mL of Hz0 to flow from syringe 6 to a syringe 8. Net yields were 5&70% in 25-30 min. Analysis
and quality assurance
Samples of [‘“F]FM were analyzed by gas chromatography using a Poropak Q column at l4O”C, and a thermal conductivity detector. “F peaks were
target
Fig.
1. Schematic diagram of manual/remote apparatus for production
of
[ ‘“F]FM.
Improved
synthesis of [‘*F]fluoromethane
detected by passing the output gas stream through a loop of tubing in a NaI(T1) well counter. Gaseous samples of [18F]FM were analyzed after dissolution by shaking with toluene; aqueous samples were injected directly into the chromatograph. Aqueous solutions of [‘*F]FM were also analyzed by HPLC, using a Cu column (25 x 0.46 cm) with H,O (1 mL/min) as mobile phase, and radioactivity and u.v.ry detection. FM eluted at 5.8 min and CH,I at 19.6 min. Results Data from optimization experiments on some of the reaction variables are shown in Fig. 2. For 20 mg of Dowex resin, 3 mL of CH3CN/(Bu)4N+ OH- was sufficient to remove >90% of the r8F (Panel A). With the reaction tube held at lOO”C, 100 PL of CH,I, 50 mg of Ag,O and a reaction time of 10 min gave
A. Elution of F_18 from co lump,
OLr 0
1
2
TBA.OH/ecetonilrile
1051
near-optimum yields (Panels B, C). In one experiment (Panel B) where the ‘*F-solution eluted from the column was evaporated to remove water before addition of CH, CN/CH, I, the speed of the reaction was increased; yields were markedly reduced by further addition of water (Panel C). Panel D shows data for cryogenic trapping of [18F]FM. In ethanol, carrier added [ lBF]FM was trapped more effectively than the NCA product, but temperatures close to the freezing point of ethanol (liquid nitrogen/methanol slush bath) were required. The SEP-PAK was effective using ethanol/dry ice (- 72°C). In experiments where [“F]FM was obtained directly from the gas phase above carrier added reaction mixtures, gas chromatograms (not shown) contained large mass peaks corresponding to FM and to CH,I. Smaller peaks corresponding to C2H50H, CH,CN and CH30H were also evident. Impurity levels for NCA [ “F]FM recovered directly from the
ion
B.
3
2
0
4
(mL) -_,
6
Time (min)
6
10
-+
D. Tratdng,
C. Reaction Darameters.
1 I 1
0
20
40
Addition
60
(mg)
60
---,
100
-120
-100
-60
Temp
-60
-40
(“C) +
Fig. 2. Optimization experiments for [ ‘*F]FM synthesis. (A) Recovery of r8F from Dowex 1 column. (B) Temporal progress of reaction using the ‘*F solution eluted in (Bu),N+/CH,CN from the Dowex column (lower curve). The upper curve, for comparison, shows data from an experiment where the “wet” ‘*F solution was evaporated to dryness and redissolved in dry CH.,CN. (C) Effects of varying Ag,O, Ag,O, CH,I and added H,O, respectively, on yield of FM (reaction ttme = 7 min). (D) Trapping of [ ‘*F]FM at various temperatures: CA, NCA refer to carrier added or no carrier added [ ‘*F]FM trapped in ethanol.
1052
S. JOHN GATLEY et al Table
I. Levels and toxicltles of non-radioactwe imourities in I ‘*FIFM “reuarations Toxicity data Amount in FM preparation
(pmol)
Impurity
Unpurified (n = 4)”
Purified (n = 6)b
OSHA limit’ (r”wL)
CH,F C,H,OH CH,CN CH,t CH,OH
to.1 0.3 * 0.2 32 + 7 77 + 32 l4+6
1991 Copyright
0883-2889/91 $3.00 + 0.00 0 1991 Pergamon Press plc
An Improved Synthesis of the Inert, Diffusible Blood-flow Tracer, [ ‘sF]Fluoromethane S. JOHN
GATLEY*, RODOLFO FRANCESCHINIt, DAVID J. SCHYLER and ALFRED
Department
of Chemistry,
Brookhaven
National
Laboratory,
RICHARD P. WOLF Upton,
FERRIERI,
NY 11973, U.S.A.
(Received 19 December 1990; in revised form 20 February 1991) An improved procedure for producing [ ‘*F]fluoromethane ([ ‘8F]FM) in batches of several hundred mCi is reported. ‘aF prepared by the InO (p, n) ‘sF reaction in a Hr’*O target is trapped on a small column of Dowex 1 ( x 10) resin to allow recovery of H, ‘*O. One new feature is elution of “F- from the column with 3 mL of CH,CN containing 67 nL_ 1.5 M aqueous (Bu),N+ OH-, after residual H,180 has been removed with dry CH,CN. This “F- solution reacts with CH,I in the nresence of Ae,O directlv to eive [ ‘*F]FM, which is swept out of the reaction vessel with a stream of air, from which CH;i and other vaiors are removed with a C,, SEP-PAK at room temperature. Another new technique is trapping of [ ‘*F]FM on a second SEP-PAK cooled in ethanol/dry ice. After warming the SEP-PAK to room temperature, the trapped [ 18F]FM can be recovered with either H,O or air. The improvements speed the preparation and minimize hands-on operations. The product has no detectable radiochemical impurities, and a specific activity of > 1 Ci/pmol. Non-radioactive CH,CN, CH,I and CH,OH are present at ~0.2 pmol per batch.
Introduction [ “FlFluoromethane
([ 18F]FM) has several advantages for positron emission tomographic (PET) measurement of local cerebral blood flow (LCBF) (Holden et al., 1981, 1983). These include the physical properties of ‘*F. The half-life of 110 min is long enough to permit distribution to PET centers from a regional radiochemical facility, and availability for several hours from a single batch synthesis. The relatively low /I,,, of the “F positron leads to lower tissue radiation doses and possibly better spatial resolution in future tomographs than for other /3+ emitters. In addition, the solubilities of FM in water and in fat (Gatley et al., 1981; Rosenthal and Nickles, 1985), which govern losses in expired air and incorporation into adipose tissue, are such that the agent has a short biological half-life, allowing repeat scanning (30-60min) in a single session. It is also easy to prepare, metabolically inert (Gatley et al., 1981) and can provide values of LCBF without the need for arterial blood sampling (Koeppe et al., 1985b). Fluoromethane may alternatively be radiolabeled with “C (Stone-Elander et al., 1988), and interest has *Author for correspondence. tPresent address: Instituto di Fisiologia Universita
Clinica de1 CNR, di Pisa, Via Savi 8, 56100 Pisa, Italy.
also been expressed (Mulholland et al., 1987) in FM labeled with “F, which has a 65 s half-life. The preparation of [ 18F]FM usually proceeds from [ ‘*F]fluoride which can be prepared by a variety of methods, but most commonly at present by the ‘*O(p, n)‘*F reaction with a water target (Nickles et al., 1986). [‘*F]FM has also been made using cyclotron targetry based on the “Ne(d, ~r)i*F reaction, via the reactive intermediates [ 18F]F2 (Wagner, 1984), N-[ ‘*F]fluoropyridone (Oberdorfer et al., 1988) and [‘*F]diethylaminosulfur trifluoride (Straatman and Welch, 1977). Fluoromethane labeled with either “C or “F has been used in basic medical research (Celesia et al., 1982; Sollevi et al., 1987) clinical applications (Celesia et al., 1984; Lagreze et al., 1987; Lagreze et al., 1988; Levine et al., 1988a,b,c, 1989; Rao et al., 1984) and in PET methodological studies (Heiss et al., 1990; Herholz and Patlak, 1987; Herholz et nl., 1989; Koeppe et al., 1985a,b; Sollevi et al., 1987). The FM method depends on tracer kinetic modeling, on a pixel-by-pixel basis, of the time-course of “F in the brain determined by a temporal sequence of scans (Holden et al., 1981). Because FM is cleared rapidly by exhalation, the biological half-life is quite short; only 2-3% of radioactivity remains in the body after 20 min. Fluoromethane can be given to most subjects by inhalation from an air-bag, followed by a brief
1049
1050
S.
JOHN
GATLEY
(30-45 s) breath hold. Alternatively it can be introduced into a spirometer and administered under a re-breathing protocol. FM can also be easily dissolved in saline and given by i.v. injection. In this case more than half of the administered dose will be expired in the first minute. Previously (Gatley et al., 1981), [18F]FM was obtained by drawing off the gas space through the septum of a sealed reaction vessel into one or more 50 mL syringes. Besides giving a rather large hand dose, this typically resulted in 2&50mL volumes in large syringes which were difficult to shield for transport. There were also concerns about chemical impurities in these preparations. The aims of the present study were to design a system for safely producing high levels (several hundred mCi) of [ ‘*F]FM for human use (either by inhalation or i.v. injection), and to characterize non-radioactive impurities in doses of [ ‘*F]FM.
Experimental Materials Vacutainer”s were obtained from Becton-Dickinson, Rutherford, NJ. Dowex 1 x 10 (200-400 mesh, Cl- form) was obtained from Becker (Phillipsburg, NJ) and converted to the OH -form as described previously (Schlyer et al., 1990). C,,SEP-PAK cartridges were purchased from Waters (Milford, MA). Preparation
of’ F-18 FM
For analysis of impurity vapors product, FM was harvested from the the reaction vessel in a 50mL syringe previously (Gatley et al., 1981), and 1 mL water. The new preparation of [“F]FM is 10 mL Vacutainer initially containing
in crude lBF head-space of as described washed with as follows. A 50mg Ag,O
et al.
and held in an oil bath at 100°C was used as a reaction vessel. The rest of the apparatus (Fig. 1), in addition to a small glass column of Dowex 1, included a 7 mL Vacutainer acting as a cold trap and a SEP-PAK to remove residual vapors. A second SEP-PAK, previously activated with 5 mL ethanol, was immersed in an ethanol/dry ice bath in order to trap [18F]FM. The components were connected with polypropylene tubing and Luer fittings, and 20 ga needles. In some initial experiments the SEP-PAKs were replaced by a 7 mL Vacutainer in which the gas stream was passed through 1 mL of ethanol at various temperatures. With valve A open, irradiated H,lsO containing 18F was pumped from the cyclotron target into the head space above the Dowex column. The Hz ‘“0 was recovered for re-use (syringe 3). Valve A was closed and valve D switched to allow 2 mL of dry CH,CN from syringe I to rinse the column (collected in syringe 4). Valves B and D were than switched to allow introduction of 3 mL of CH,CN containing 67pL of 1.5 M aqueous (Bu),N+ OHinto the reaction vessel. A 1 mL sample of CH,CN containing 100 p L of CH, I were then similarly added (syringe 5), after switching valves B and E appropriately. The reaction mixture was heated for lOmin, and then 50mL air was drawn through the system (syringe 7) at a rate which caused the solvents to boil gently. Trapped [ ‘*F]FM was recovered, after allowing SEP-PAK 2 to warm to room temperature, by switching valve C to allow 5 mL of Hz0 to flow from syringe 6 to a syringe 8. Net yields were 5&70% in 25-30 min. Analysis
and quality assurance
Samples of [‘“F]FM were analyzed by gas chromatography using a Poropak Q column at l4O”C, and a thermal conductivity detector. “F peaks were
target
Fig.
1. Schematic diagram of manual/remote apparatus for production
of
[ ‘“F]FM.
Improved
synthesis of [‘*F]fluoromethane
detected by passing the output gas stream through a loop of tubing in a NaI(T1) well counter. Gaseous samples of [18F]FM were analyzed after dissolution by shaking with toluene; aqueous samples were injected directly into the chromatograph. Aqueous solutions of [‘*F]FM were also analyzed by HPLC, using a Cu column (25 x 0.46 cm) with H,O (1 mL/min) as mobile phase, and radioactivity and u.v.ry detection. FM eluted at 5.8 min and CH,I at 19.6 min. Results Data from optimization experiments on some of the reaction variables are shown in Fig. 2. For 20 mg of Dowex resin, 3 mL of CH3CN/(Bu)4N+ OH- was sufficient to remove >90% of the r8F (Panel A). With the reaction tube held at lOO”C, 100 PL of CH,I, 50 mg of Ag,O and a reaction time of 10 min gave
A. Elution of F_18 from co lump,
OLr 0
1
2
TBA.OH/ecetonilrile
1051
near-optimum yields (Panels B, C). In one experiment (Panel B) where the ‘*F-solution eluted from the column was evaporated to remove water before addition of CH, CN/CH, I, the speed of the reaction was increased; yields were markedly reduced by further addition of water (Panel C). Panel D shows data for cryogenic trapping of [18F]FM. In ethanol, carrier added [ lBF]FM was trapped more effectively than the NCA product, but temperatures close to the freezing point of ethanol (liquid nitrogen/methanol slush bath) were required. The SEP-PAK was effective using ethanol/dry ice (- 72°C). In experiments where [“F]FM was obtained directly from the gas phase above carrier added reaction mixtures, gas chromatograms (not shown) contained large mass peaks corresponding to FM and to CH,I. Smaller peaks corresponding to C2H50H, CH,CN and CH30H were also evident. Impurity levels for NCA [ “F]FM recovered directly from the
ion
B.
3
2
0
4
(mL) -_,
6
Time (min)
6
10
-+
D. Tratdng,
C. Reaction Darameters.
1 I 1
0
20
40
Addition
60
(mg)
60
---,
100
-120
-100
-60
Temp
-60
-40
(“C) +
Fig. 2. Optimization experiments for [ ‘*F]FM synthesis. (A) Recovery of r8F from Dowex 1 column. (B) Temporal progress of reaction using the ‘*F solution eluted in (Bu),N+/CH,CN from the Dowex column (lower curve). The upper curve, for comparison, shows data from an experiment where the “wet” ‘*F solution was evaporated to dryness and redissolved in dry CH.,CN. (C) Effects of varying Ag,O, Ag,O, CH,I and added H,O, respectively, on yield of FM (reaction ttme = 7 min). (D) Trapping of [ ‘*F]FM at various temperatures: CA, NCA refer to carrier added or no carrier added [ ‘*F]FM trapped in ethanol.
1052
S. JOHN GATLEY et al Table
I. Levels and toxicltles of non-radioactwe imourities in I ‘*FIFM “reuarations Toxicity data Amount in FM preparation
(pmol)
Impurity
Unpurified (n = 4)”
Purified (n = 6)b
OSHA limit’ (r”wL)
CH,F C,H,OH CH,CN CH,t CH,OH
to.1 0.3 * 0.2 32 + 7 77 + 32 l4+6
